• 한국통신학회논문지
• /
• 제21권8호
• /
• pp.1874-1884
• /
• 1996

In this paper, we proposed automatic searching algorithm of thresholding value using multilevel thresholding for face segmentation from input bust image effectively. The proposed algorithm extracted the thresholding value of brightness that is formed background region, face region and hair region without illumination, background and face size from input image. The statistical projection analysis project the brightness of multilevel thresholding image into horizontal and vertical direction and decide the thresholding value of face. And the algorithm extracted elliptical type block of face from input image in order to reduce the back ground region and hair region efficiently. The proposed algorithm can reduce searching area of feature extraction and processing time for face recognication.

• Journal of Information Processing Systems
• /
• 제13권4호
• /
• pp.677-688
• /
• 2017

In order to solve the undetected probability of multiple targets in ultra-wideband (UWB) through-the-wall radar imaging (TWRI), a time-delay and amplitude modified back projection (BP) algorithm is proposed. The refraction point is found by Fermat's principle in the presence of a wall, and the time-delay is correctly compensated. On this basis, transmission loss of the electromagnetic wave, the absorption loss of the refraction wave, and the diffusion loss of the spherical wave are analyzed in detail. Amplitude compensation is deduced and tested on a model with a single-layer wall. The simulating results by finite difference time domain (FDTD) show that it is effective in increasing the scattering intensity of the targets behind the wall. Compensation for the diffusion loss in the spherical wave also plays a main role. Additionally, the two-layer wall model is simulated. Then, the calculating time and the imaging quality are compared between a single-layer wall model and a two-layer wall model. The results illustrate the performance of the time-delay and amplitude-modified BP algorithm with multiple targets and multiple-layer walls of UWB TWRI.

• 한국의학물리학회지:의학물리
• /
• 제14권3호
• /
• pp.167-174
• /
• 2003

정위적방사선수술과 같은 경우 치료계획 수립 시 병변의 정확한 위치뿐만 아니라 정확한 부피와 모양을 아는 것도 매우 중요하다. 병변의 확인을 위해서 때로는 혈관조영영상이 이용되기도 하는데 동정맥 기형과 같은 경우 이 방법이 병변의 구별을 위하여 가장 좋은 방법이기 때문이다. 병변의 정확한 위치는 두개의 투사영상으로부터 얻을 수 있지만 두 개의 투사영상 만으로는 병변을 3차원적으로 재구성하는 것은 불가능하다고 여겨지고 있다. 본 연구의 목적은 다수의 투사 영상들을 이용하여 병변을 3차원적으로 재구성하는 것이다. 이때 병변의 위치는 기존에 제안된 방법에 의하여 이미 알고 있다고 가정하였으며 모든 과정은 병변의 중심을 원점으로 하는 표적좌표계에서 수행되었다. 본 연구에서는 6개의 투사영상이 이용되었는데 정면과 측면 투사영상은 체적소(voxel)로 구성된 재구성상자를 구하기 위하여 이용되었으며 나머지 네 개의 투사영상은 역투사 방법(back-projection method)에 의하여 재구성 상자(Reconstruction Box) 내에서 3차원적으로 재구성하는데 이용되었다. 이 방법의 정확도와 해상도는 병변의 크기와 모양에 따라 달라질 수 있다. 본 연구에서 제안된 알고리듬의 검증을 위하여 C 언어와 Matlab을 이용하여 타원체 모델과 말굽형 모델에 대하여 투사영상을 얻고 그 영상을 이용하여 재구성해보았다. 타원체모델의 경우에는 원래의 모델보다 약간 크게 재구성되었지만 모양과 방향, 위치가 정확함을 확인할 수 있었다. 말굽형 모델은 재구성된 모양이 원래의 모양과 차이가 많이 났지만 기존 방법에 비하여 실제 모양에 근접하게 재구성할 수 있었으므로 병변을 확인하는 경우에는 도움이 될 것으로 사료된다.

• 한국전자파학회논문지
• /
• 제29권9호
• /
• pp.722-730
• /
• 2018

본 논문에서는 FMCW(frequency modulated continuous wave) 신호기반 SAR 시스템의 스퀸트 각도(squint angle) 설정에 따른 복원영상 내 목표물 분산전력패턴 변화를 예측하기 위한 분석기법에 대해 설명한다. SAR 복원영상 내 목표물 분산 전력패턴은 지면에 투영된 전파진행방향과 합성개구면과 스퀸트 각도가 이루는 방위방향에 따라 일정한 방향으로 변형된 특성을 갖게 되며, 이와 같은 특성변화를 3차원 SAR 기하구조 내에서 분석한다. 스퀸트 설정각도에 따라 변형된 목표물 분산전력패턴은 FMCW 신호모델과 back-projection 알고리즘을 기반으로 한 모의실험을 통해 그 분석결과를 검증한다. 또한, 변형된 분산전력패턴의 예측결과는 SAR 기하구조와 스퀸트 설정각도를 변수로 정리된다.

• 전자공학회논문지B
• /
• 제33B권9호
• /
• pp.49-61
• /
• 1996

Back-propagation(BP) algorithm needs a lot of time to train the artificial neural network (ANN) to get high accuracy level in classification tasks. So there have been extensive researches to process back-propagation algorithm on parallel processors. This paper prsents a linear systolic array which calculates forward-backward propagation of BP algorithm at the same time using effective space-time transformation and PE structure. First, we analyze data flow of forwared and backward propagations and then, represent the BP algorithm into data dapendency graph (DG) which shows parallelism inherent in the BP algorithm. Next, apply space-time transformation on the DG of ANN is turn with orthogonal direction projection. By doing so, we can get a snakelike systolic array. Also we calculate the interval of input for parallel processing, calculate the indices to make the right datas be used at the right PE when forward and bvackward propagations are processed in the same PE. And then verify the correctness of output when forward and backward propagations are executed at the same time. By doing so, the proposed system maximizes parallelism of BP algorithm, minimizes th enumber of PEs. And it reduces the execution time by 2 times through making idle PEs participate in forward-backward propagation at the same time.

• 반도체디스플레이기술학회지
• /
• 제11권4호
• /
• pp.49-55
• /
• 2012

This paper proposes an efficient method of object identification and localization for image recognition. The new proposed algorithm utilizes correlogram back-projection in the YCbCr chromaticity components to handle the problem of sub-region querying. Utilizing similar spatial color information enables users to detect and locate primary location and candidate regions accurately, without the need for additional information about the number of objects. Comparing this proposed algorithm to existing methods, experimental results show that improvement of 21% was observed. These results reveal that color correlogram is markedly more effective than color histogram for this task. Main contribution of this paper is that a different way of treating color spaces and a histogram measure, which involves information on spatial color, are applied in object localization. This approach opens up new opportunities for object detection for the use in the area of interactive image and 2-D based augmented reality.

• Korean Journal of Radiology
• /
• 제23권7호
• /
• pp.752-762
• /
• 2022

Objective: To compare a deep learning-based reconstruction (DLR) algorithm for pediatric abdominopelvic computed tomography (CT) with filtered back projection (FBP) and iterative reconstruction (IR) algorithms. Materials and Methods: Post-contrast abdominopelvic CT scans obtained from 120 pediatric patients (mean age ± standard deviation, 8.7 ± 5.2 years; 60 males) between May 2020 and October 2020 were evaluated in this retrospective study. Images were reconstructed using FBP, a hybrid IR algorithm (ASiR-V) with blending factors of 50% and 100% (AV50 and AV100, respectively), and a DLR algorithm (TrueFidelity) with three strength levels (low, medium, and high). Noise power spectrum (NPS) and edge rise distance (ERD) were used to evaluate noise characteristics and spatial resolution, respectively. Image noise, edge definition, overall image quality, lesion detectability and conspicuity, and artifacts were qualitatively scored by two pediatric radiologists, and the scores of the two reviewers were averaged. A repeated-measures analysis of variance followed by the Bonferroni post-hoc test was used to compare NPS and ERD among the six reconstruction methods. The Friedman rank sum test followed by the Nemenyi-Wilcoxon-Wilcox all-pairs test was used to compare the results of the qualitative visual analysis among the six reconstruction methods. Results: The NPS noise magnitude of AV100 was significantly lower than that of the DLR, whereas the NPS peak of AV100 was significantly higher than that of the high- and medium-strength DLR (p < 0.001). The NPS average spatial frequencies were higher for DLR than for ASiR-V (p < 0.001). ERD was shorter with DLR than with ASiR-V and FBP (p < 0.001). Qualitative visual analysis revealed better overall image quality with high-strength DLR than with ASiR-V (p < 0.001). Conclusion: For pediatric abdominopelvic CT, the DLR algorithm may provide improved noise characteristics and better spatial resolution than the hybrid IR algorithm.

• 한국산학기술학회논문지
• /
• 제13권11호
• /
• pp.5311-5316
• /
• 2012

뇌 영상분석 알고리즘 적용에 따른 뇌혈류의 변화를 관찰하기 위해 정상 성인 13명(평균연령 39세)을 대상으로 뇌혈류 단 광자 단층촬영(Single Photon Emission Computed Tomography, SPECT)을 시행하였다. 획득된 영상을 여과후 역투영법(Filtered Back Projection, FBP)과 반복적 방법(Ordered Subset Expectation Maximization, OSEM)으로 영상을 재구성하여 통계학적 파라미터 뇌지도법으로 비교하여 방사성 의약품의 뇌 분포양상을 확인하였으며, 혈류의 변화는 크러스트(Cluster)로 표현시켰다. 이에 대한 결과로는 여과후 역투영법이 반복적 방법보다 섭취가 증가된 부위는 우측 전두엽, 대뇌회전하부, 외핵, 좌측대뇌 변엽과 대상이랑이며, 섭취가 감소된 부위는 좌측전두엽, 중간전두이랑, 하측전두이랑, 중심전이랑, 하측전두이랑, 중심전이랑이었다. 이는 영상재구성시 적용되는 알고리즘에 따라 뇌 혈류분포가 다르게 나타난 것을 크러스트(Cluster)로 표현시켰고, 명확한 시각적 표시가 가능하도록 뇌 확룔 지도로 보여주는데 의의가 있다.

• 제어로봇시스템학회논문지
• /
• 제14권8호
• /
• pp.818-823
• /
• 2008

In this paper, we propose a feature extraction method which extracts directional features of handwritten numerals by using the projection runlength. Our directional featrures are obtained from four directional images, each of which contains horizontal, vertical, right-diagonal and left-diagonal lines in entire numeral shape respectively. A conventional method which extracts directional features by using Kirsch masks generates edge-shaped double line directional images for four directions, whereas our method uses the projections and their runlengths for four directions to produces single line directional images for four directions. To obtain the directional projections for four directions from a numeral image, some preprocessing steps such as thinning and dilation are required, but the shapes of resultant directional lines are more similar to the numeral lines of input numerals. Four [$4{\times}4$] directional features of a numeral are obtained from four directional line images through a zoning method. By using a hybrid feature which is made by combining our feature with the conventional features of a mesh features, a kirsch directional feature and a concavity feature, higher recognition rates of the handwrittern numerals can be obtained. For recognition test with given features, we use a multi-layer perceptron neural network classifier which is trained with the back propagation algorithm. Through the experiments with the handwritten numeral database of Concordia University, we have achieved a recognition rate of 97.85%.

• 한국위성정보통신학회논문지
• /
• 제8권1호
• /
• pp.77-82
• /
• 2013

최근 주목을 받고 있는 Particle Filtering은 실제 객체 추적에서 발생하는 비선형, 비 가우시안 분포를 가지는 상태 벡터의 사후확률을 추정하기 위한 Monte Carlo 시뮬레이션에 기반을 둔 추적 방법론이다. 우리는 본 논문에서 Particle Filtering을 이용한 객체 추적성능을 향상시킬 수 있는 두 가지 방법론을 제안한다. 첫 번째는 확률이 가장 낮은 샘플을 이전 프레임의 추정된 상태 벡터로 대치하는 피드백 방법론이고, 두 번째는 객체 확률 분포를 추정된 객체 후보영역에 역투영하여 신뢰구간을 구함으로써 추적 박스의 정확도를 향상시키는 방법이다. 또한, 실험을 통해 구한 추적 샘플의 진화 방정식을 제시하였다. 우리는 다양한 상황이 설정된 실험 데이터 셋을 구성하여 실험을 실시하여 제안한 방법론의 우수성을 입증하였다.